JP5567044B2 - Wind farm operation method and wind farm operation control system - Google Patents

Description

  The present invention relates to a wind farm operation method and a wind farm operation control system, and more particularly to a wind farm operation method and a wind farm operation control system capable of being operated with excellent economic efficiency.

In recent years, wind power generators that generate power using wind as a renewable energy have been widely used from the viewpoint of conservation of the global environment. A wind turbine generator generally has a rotor having a plurality of blades attached to a hub. The rotor is mounted on a nacelle located on a tower erected on land or offshore. In a wind turbine generator, the blade receives wind and the rotor rotates, and the rotation of the rotor is transmitted to the generator housed in the nacelle via the gearbox, and electric power is generated in the generator. Yes.
Examples of the speed increaser provided between the rotor and the generator include a gear type equipped with a planetary gear and a hydraulic transmission type combined with a hydraulic pump and a hydraulic motor.

  A wind turbine generator is made up of a number of components such as blades, gears and hydraulic transmissions, hubs, nacelles, various bearings such as main shaft bearings and yaw bearings, and towers. Composed. If these parts deteriorate or fail, the wind turbine generator may be forced to stop operating. Therefore, it is important to monitor the state of the components of the wind power generator to prevent an unexpected stop of the operation of the wind power generator.

  Therefore, Patent Document 1 describes the operation of a wind power generator that predicts the remaining life of a component by a state monitoring system (CMS) and limits the output (power generation amount) so that the expected remaining life is obtained. A method is disclosed. For example, if it is predicted that the remaining life of a part will be exhausted by the time of the next periodic maintenance, the output is lowered to extend the remaining life of the part until the next periodic maintenance time. As a result, in addition to preventing an unexpected shutdown due to a component failure, it is possible to reduce the need for unscheduled maintenance and increase the profit obtained from the wind turbine generator.

US Patent Application Publication No. 2010/0332272

By the way, in wind farms, the damage or deterioration of parts varies from wind turbine generator to wind turbine generator, so the operating conditions of each wind turbine generator that maximizes the profits from each wind turbine generator maximize the profits of the entire wind farm. It does not always match the operating conditions.
For example, in a wind farm composed of a first windmill and a second windmill, the speed of deterioration of parts of each windmill often varies. Therefore, if we pursue the maximization of profits obtained from each wind turbine, the output of each wind turbine will be extended so that the remaining life of the parts can be extended until the optimum maintenance time (maintenance time when the profit of each wind turbine is maximized). Limited. However, in the same wind farm, maintenance costs may be reduced by performing maintenance on the parts of the first windmill and the second windmill at the same time. In particular, wind farms installed in difficult-to-access locations (offshore and mountainous areas) can greatly reduce costs by maintaining multiple wind turbines at the same time. For this reason, the maintenance time of the wind turbine with the slower component deterioration (first wind turbine) is set ahead of the optimum maintenance time at which the profit of the wind turbine is maximized, and the wind turbine with the faster component deterioration (second wind turbine) At the same time, maintenance may increase the profits earned from the entire wind farm. On the other hand, if the effect of changing the maintenance period on the power sales revenue of each windmill is neglected because the cost reduction effect due to the simultaneous maintenance of the first windmill and the second windmill will be obtained, the profit of the wind farm as a whole will be rejected. It can be detrimental.

  In this respect, the operation method described in Patent Document 1 is intended for a single wind power generator, and is not intended to improve the profitability of a wind farm including a plurality of wind power generators. Therefore, even if the application target of the operation method described in Patent Document 1 is simply expanded from the wind turbine generator alone to the entire wind farm, it is not possible to expect a significant increase in the profit of the entire wind farm.

  The present invention has been made in view of the above-described circumstances, and an object thereof is to provide a wind farm operation method and a wind farm operation control system that can effectively improve the profit of the entire wind farm. .

  A wind farm operation method according to the present invention is a wind farm operation method including a plurality of wind turbines, the step of predicting the remaining life of parts for each wind turbine, and a plurality of output restriction conditions for each wind turbine. Predicting power sales revenue, predicting maintenance costs under each output restriction condition based on the remaining life of the parts for each windmill, and at least predicting each windmill for each output restriction condition Based on the power sales revenue and the maintenance cost, the step of selecting an output restriction condition for each wind turbine that maximizes the profit obtained from the wind farm, and the operation control of each wind turbine based on the selected output restriction condition And performing.

  According to this wind farm operation method, power sales revenue and maintenance costs are predicted for each output limit condition for each wind farm, and the output limit condition that maximizes the profit obtained from the wind farm is selected for each wind turbine. As a result, the profit of the entire wind farm can be improved effectively. That is, instead of paying attention to the profit from each windmill, the operating condition of each windmill is determined by paying attention to the profit of the entire wind farm, thereby making it possible to operate the wind farm with excellent economic efficiency.

In the wind farm operating method, the maintenance cost is predicted for each of a plurality of candidate times for the next maintenance of the component, and at least the output restriction condition and the sales price predicted for each candidate time for each wind turbine. Based on the electricity revenue and the maintenance cost, an output restriction condition and a maintenance time at which the profit obtained from the wind farm is maximized may be determined for each wind turbine.
As a result, it is possible to select the combination that is most expected to increase the profitability of the entire wind farm from among a plurality of combinations of the output restriction conditions and the maintenance time for each wind turbine, and the wind farm operation that is more economical Is possible.

In the wind farm operating method, in the step of predicting the remaining life, the remaining life of the component may be predicted based on a change in a state value indicating a damaged state or a deteriorated state of the component.
Thereby, the state value indicating the damaged state or the deteriorated state of the component can be monitored, and the remaining life of the component can be predicted with high accuracy.
The state value indicating the damage state or deterioration state of the component can include, for example, blade weight, weight unbalance between blades, blade vibration, etc. as an indication of blade damage state or deterioration state. As an example of the damage state or deterioration state of the tower, fatigue load of the tower base or the upper part of the tower, vibration of the tower, etc. can be mentioned. As an indication of the damage state or deterioration state of the gearbox, the main shaft bearing or hydraulic pump bearing Vibration, main shaft vibration and swing, hydraulic pump piston vibration and amplitude, hydraulic transmission efficiency, etc., which can indicate damage or deterioration of casting parts such as nacelle base plate and hub (rotor head) As an example, fatigue due to stress concentration can be mentioned.

The wind farm operating method may further include a step of stopping the operation of the wind turbine having the component when the state value of the component is out of a management range.
As a result, when the state value of the component is out of the control range before the complete failure of the component, the operation of the wind turbine having the component is stopped, and the wind turbine is forced to stop for a long time. Can be prevented.

  The wind farm operation control system according to the present invention is a wind farm operation control system including a plurality of wind turbines, and includes a remaining life prediction unit that predicts the remaining life of components for each wind turbine, and a plurality of wind turbine operations for each wind turbine. A power sales revenue prediction unit that predicts power sales revenue under the output restriction conditions, and a maintenance cost prediction unit that predicts maintenance costs under each output restriction condition based on the remaining life of the parts for each wind turbine, at least, An output restriction condition selection unit that selects, for each windmill, an output restriction condition that maximizes the profit obtained from the wind farm based on the power sales revenue and the maintenance cost predicted for each of the output restriction conditions for each windmill; A driving command unit that sends a driving command to each wind turbine based on the selected output restriction condition. And features.

  According to the wind farm operation control system, the power sales revenue and maintenance cost are predicted for each wind turbine in each output limit condition, and the output limit condition that maximizes the profit obtained from the wind farm is selected for each wind turbine. As a result, the profit of the entire wind farm can be improved effectively. That is, instead of paying attention to the profit from each windmill, the operating condition of each windmill is determined by paying attention to the profit of the entire wind farm, thereby making it possible to operate the wind farm with excellent economic efficiency.

  According to the present invention, the power sales revenue and the maintenance cost are predicted for each wind turbine of the wind farm for each output limitation condition, and the output limitation condition that maximizes the profit obtained from the wind farm is selected for each wind turbine. This can effectively improve the profitability of the whole wind farm. That is, instead of paying attention to the profit from each windmill, the operating condition of each windmill is determined by paying attention to the profit of the entire wind farm, thereby making it possible to operate the wind farm with excellent economic efficiency.

It is a figure which shows the structural example of the operation control system of a wind farm. It is a figure which shows an example of the windmill which comprises a wind farm. It is a figure which shows the other example of the windmill which comprises a wind farm. It is an example of the graph which shows the time-dependent change of the state value which shows the damage state or deterioration state of components. It is a graph which shows the time-dependent change of the maintenance cost of components under each output restriction condition. It is a table | surface which shows an example of the maintenance cost C estimated for every output restriction condition and maintenance candidate time. It is a table | surface which shows an example of the maintenance cost which considered cost reduction effect (DELTA) C by simultaneous maintenance with another windmill. It is a flowchart which shows the procedure of the operating method of a wind farm. It is the figure which showed a mode that the output restriction conditions and the maintenance time of the windmill A and the windmill B were determined in embodiment of this invention. It is the figure which showed a mode that the output restriction conditions and the maintenance time of the windmills A and B were determined when the maximization of the profit of each windmill was pursued.

  Hereinafter, embodiments of the present invention will be described with reference to the accompanying drawings. However, the dimensions, materials, shapes, relative arrangements, and the like of the components described in this embodiment are not intended to limit the scope of the present invention unless otherwise specified, and are merely illustrative examples. Only.

  FIG. 1 is a diagram illustrating a configuration example of an operation control system for a wind farm. FIG. 2 is a view showing an example of a windmill constituting a wind farm. FIG. 3 is a view showing another example of a wind turbine constituting a wind farm.

  An operation control system 10 illustrated in FIG. 1 manages operation control of a wind farm 1 including a plurality of windmills 100A to 100C. Hereinafter, after giving examples of the windmills 100A to 100C constituting the wind farm 1, details of the operation control system 10 will be described.

  As shown in FIG. 2, each windmill 100 </ b> A to 100 </ b> C of the wind farm 1 includes a rotor 105 including a plurality of blades 102 and a hub 104 to which the blades 102 are attached, a main shaft 106 coupled to the hub 104, and a main shaft 106. There may be provided a gear-type gearbox 108 that speeds up the rotation of the motor and a generator 110 connected to the output shaft 109 of the gearbox 108. Further, the speed increaser 108 and the generator 110 may be accommodated in a nacelle 112 that rotatably supports the main shaft 106 via a main shaft bearing 107. The nacelle base plate 114 constituting the bottom surface of the nacelle 112 may be supported by the tower 116 via a yaw slewing bearing 118. The nacelle base plate 114 is fixed with a yaw turning mechanism 119 having a yaw motor and a pinion gear. The yaw motor is engaged with the ring gear provided on the tower 116 side in mesh with the pinion gear of the yaw turning mechanism 119. The nacelle 112 may be turned with respect to the tower 116 by being driven. Further, each blade 102 is supported by the hub 104 via a blade slewing bearing (not shown), and the pitch angle is adjusted by a pitch drive actuator (not shown) provided in the hub 104. May be.

In addition, each windmill 100A to 100C of the wind farm 1 is replaced with a drive train provided with a gear-type gearbox 108, and as shown in FIG. The rotational energy may be transmitted to the generator 110.
As shown in FIG. 3, the hydraulic transmission 120 rotates with the main shaft 106 to generate pressure oil, and a hydraulic motor 124 that is driven by the pressure oil from the hydraulic pump 122 and inputs rotation to the generator 110. With. The outlet of the hydraulic pump 122 is connected to the inlet of the hydraulic motor 124 by a high pressure oil line 126. On the other hand, the inlet of the hydraulic pump 122 is connected to the outlet of the hydraulic motor 124 by a low-pressure oil line 128. As a result, the hydraulic pump 122 driven by the rotation of the main shaft 106 pressurizes the hydraulic oil supplied from the low pressure oil line 128 and discharges it to the high pressure oil line 126, and is driven by the pressurized hydraulic oil (pressure oil). Rotational energy is input from the hydraulic motor 124 to the generator 110. In this way, the rotational energy of the main shaft 106 is transmitted to the generator 110 via the hydraulic transmission 120.

In the wind turbine 100 having the configuration example shown in FIG. 2 or 3, a state value X indicating a damaged state or a deteriorated state of various parts is acquired by a state value detection sensor (not shown), and the state monitoring unit 12 of the operation control system 10. The status value X is reported in
Specific examples of the state values X of the various components of the wind turbines 100A to 100C include the weight of the blade 2, the weight unbalance between the blades 102, the vibration of the blade 102, and the like as indicating the damaged state or the deteriorated state of the blade 102. Examples of the damaged or degraded state of the tower 116 include fatigue load at the base of the tower 116 or the upper portion of the tower 116, vibration of the tower 116, and the like. Examples of the deterioration state include vibration of the main shaft bearing 107 and the hydraulic pump 122, vibration and swing of the main shaft 106, piston vibration and amplitude of the hydraulic pump 122, efficiency of the hydraulic transmission 120, and the like. Damaged or deteriorated state of members made of castings such as 114 and hub 104 Mention may be made of the fatigue due to the stress concentration as shown.

For example, an optical fiber sensor (FBG: Fiber Bragg Grating) embedded in each blade 102 is used as the state value detection sensor, and distortion measured by the optical fiber sensor is converted into a bending moment, and a sine wave included in the bending moment is obtained. The weight of the blade 102, which is an example of the state value X, may be obtained from the function component. As a factor of the weight change of the blade 102, it is conceivable that the blade 102 is damaged due to a lightning strike or a collision with a flying object, or the blade 102 is iced. Further, the weight imbalance (moment of the rotor 105) of the blade 102, which is an example of the state value X, may be obtained from the weight of each blade 102 obtained by the same method. Further, an acceleration sensor attached to the blade 102 may be used as the state value detection sensor, and vibration of the blade 102, which is an example of the state value X, may be obtained from the measurement result of the acceleration sensor.
Alternatively, a strain gauge attached to the upper portion of the tower 116 may be used as the state value detection sensor, and a fatigue load on the upper portion of the tower 116 that is an example of the state value X may be obtained from bending or twisting measured by the strain gauge. . Further, a strain gauge attached to the lower portion of the tower 116 may be used as the state value detection sensor, and the fatigue load at the base of the tower 116, which is an example of the state value X, may be obtained from the bending or load measured by the strain gauge. . Furthermore, an acceleration sensor attached to the tower 116 may be used as the state value detection sensor, and the vibration of the tower 116 as an example of the state value X may be obtained from the measurement result of the acceleration sensor.
Alternatively, when the vibrations of various bearings represented by the main shaft bearing 107, the wing slewing bearing, the yaw slewing bearing 118, the hydraulic pump 122 and the bearing provided in the hydraulic motor 124 (pump bearing and motor bearing) are the state value X, An acceleration sensor attached directly or indirectly to these bearings may be used as the state value detection sensor. Further, when the vibration or swing of the main shaft 106 is the state value X, an acceleration sensor attached to the main shaft 106 may be used as the state value detection sensor. When the hydraulic pump 122 and the hydraulic motor 124 include a plurality of pistons, a cylinder that guides the pistons, and a cam that moves each piston up and down or a cam that rotates when the piston moves up and down, An acceleration sensor attached to the casing of the hydraulic motor 124 may be used as the state value detection sensor, and the vibration / amplitude of the piston as the state value X may be obtained. Further, when the efficiency of the hydraulic transmission 120 is the state value X, the output of the hydraulic pump 122 is calculated from the measured values of these sensors using a pressure sensor installed in the high-pressure oil line 126 and a rotation speed meter attached to the main shaft 106. The efficiency of the hydraulic transmission 120 may be obtained by calculating and dividing the electrical output of the generator 110 measured using a voltage / current sensor by the output of the hydraulic pump 122. In this case, a combination of a pressure sensor, a tachometer, and a voltage / current sensor functions as a state value detection sensor.

  The acquisition of the state value X by the state value detection sensor may be performed continuously or periodically, or may be performed when a predetermined condition is satisfied. For example, when a lightning strike to the blade 102 is detected by a lightning strike sensor (such as a current sensor installed at the blade root), the state value X of the blade 102 may be acquired by the state value detection sensor.

  The wind turbines 100 </ b> A to 100 </ b> C having the above configuration are connected to the operation control system 10 through a signal line (bus) 2 as illustrated in FIG. 1. The signal line 2 exchanges information between the windmills 100 </ b> A to 100 </ b> C and the operation control system 10. For example, the signal line 2 is used to supply information related to the state of the components of each of the windmills 100A to 100C to the operation control system 10 and to supply a control signal from the operation control system 10 to each of the windmills 100A to 100C. .

  In the example illustrated in FIG. 1, the operation control system 10 mainly includes a state monitoring unit 12, a remaining life prediction unit 14, a maintenance cost prediction unit 16, a power sales revenue prediction unit 18, a revenue estimation unit 20, and an output. A restriction condition selection unit 22, a maintenance time selection unit 23, and an operation command unit 24 are provided.

The state monitoring unit 12 receives, from the wind turbines 100A to 100C, the state value X indicating the damaged state or the deteriorated state of the components of the wind turbines 100A to 100C via the signal line 2, and the state of the components of the wind turbines 100A to 100C. To monitor. A state value indicating a damaged state or a deteriorated state of the component is acquired by the above-described state value detection sensor provided in each of the windmills 100A to 100C. The state monitoring unit 12 determines whether or not the state values of the components of the wind turbines 100A to 100C are within the management range. When the state monitoring unit 12 determines that the state values of the components of a certain windmill 100A to 100C are out of the management range, the operation of the windmill is stopped by a command from the operation command unit 24 described later. For example, the state value of the component may be stopped the operation of the wind turbine has reached a predetermined threshold value of the safety side than the threshold value X _th below corresponding to the component total loss _(see FIG. 4). On the other hand, for the wind turbines 100A to 100C in which the component state values are within the management range, the operation is continued and the remaining life prediction unit 14 predicts the remaining life of the component.

  The remaining life prediction unit 14 predicts the remaining life of components for each of the wind turbines 100A to 100C. The remaining life prediction unit 14 may predict the remaining life of the parts based on, for example, a change over time in a state value indicating a damaged state or a deteriorated state of the parts of each of the wind turbines 100A to 100C. In addition, what is preserve | saved in the state value database 15 may be used for the time-dependent change of the state value of components.

FIG. 4 is an example of a graph showing a change over time in a state value indicating a damaged state or a deteriorated state of a component. In the graph shown in the figure, the wind turbine maximum output is not limited from time t ₀ to time tc (however, the output is controlled below the rated output), and the wind turbine maximum output is 80 at the rated output at time t _c . % to be limited, at time t _M maintenance part represents the temporal change of state values when done.
According from time t ₀ to the time course 30 state value X of the current until the time t _1, until time t _{0 ~} time t _c is rapidly increased state value X, the wind turbine maximum output at time t _c is the rated output reduces the rate of increase of the state value X to which is limited to 80%, state value X for the maintenance of the parts is performed at time t _M is then lowered to the initial value, increased state value X again over time ing. Such a change 30 with time is stored in the state value database 15 and is used for the remaining life prediction of the component in the remaining life prediction unit 14.

In the remaining service life prediction unit 14, the change curve 32A~32C state value X at a future current time _{t 1} later estimated for each output limit condition, the threshold _{X th} state value X using these change curves 32A~32C The time to reach (part remaining life T _100% , T _90% , T _80% ,...) May be obtained. Threshold X _th is a state value corresponding to the total loss of the part. Here, the remaining life T _100%, meaning the remaining life of the part in the case of no restriction of the windmill maximum output (the time until the state value X reaches the threshold X _th). Similarly, the remaining life T _90% means the remaining life of the component when the maximum output of the wind turbine is limited to 90% of the rated output, and the remaining life T _80% is the case where the maximum output of the wind turbine is limited to 80% of the rated output. Means the remaining life of the part.
The remaining life predicting unit 14 is safer than the threshold X _th in addition to the time until the state value X reaches the threshold X _th (part remaining lifetime T _100% , T _90% , T _80% ,...). The time until the state value X reaches the threshold values X ₁ and X ₂ on the side may be obtained. The threshold values X ₁ and X ₂ are critical state values such that the maintenance cost rapidly changes with these values as a boundary.

  The prediction result by the remaining life prediction unit 14 is used to predict the maintenance cost of each of the wind turbines 100 </ b> A to 100 </ b> C in the maintenance cost prediction unit 16.

The maintenance cost predicting unit 16 predicts the maintenance cost under each output restriction condition for each of the wind turbines 100A to 100C based on the remaining lifetimes T _100% , T _90% , T _80% ,. The maintenance cost may be the total cost estimated for each part of each windmill 100A to 100C.

FIG. 5 is a graph showing changes over time in the maintenance cost of parts under each output limiting condition.
In the example shown in the figure, there are three types of maintenance costs C: C ₁ , C ₂ , and C _full according to the damage / degradation stages of the parts corresponding to the threshold values X ₁ , X ₂ , and X _th of the part state values. is there. Specifically, the maintenance condition value X of parts but maintenance is not maintenance costs are generated not required in the case of less than X _1, the component state value X is in the first stage the most convenient becomes the X ₁ or more There is room to do. To perform maintenance of the first stage requires maintenance costs C _1. Further progress is part of the damage or deterioration state value X of the part is the threshold value X ₂ or more, the room for results in a relatively simple maintenance of the second stage. Maintenance cost C ₂ (> C ₁ ) is required to perform the second-stage maintenance. Proceeds further component damage or deterioration, when the state value X of the part reaches the threshold X _th corresponding to the component total loss, maintenance of the final stage with a replacement or overhaul of the parts is required, the wind turbine having the components Forced to stop operation. Maintenance cost C _full (> C ₂ ) is required to perform final-stage maintenance.

  In addition, since the prediction result by the remaining life prediction unit 14 depends on each output restriction condition, the maintenance cost C predicted by the maintenance cost prediction unit 16 is different for each output restriction condition, and indicates temporal changes 34A, 34B, and 34C, respectively. The time-dependent change 34A corresponds to the case where the wind turbine maximum output is not limited, the time-dependent change 34B corresponds to the case where the wind turbine maximum output is limited to 90% of the rated output, and the time-dependent change 34C corresponds to the wind turbine maximum output. This corresponds to the case where the output is limited to 80% of the rated output.

The maintenance cost predicting unit 16 may predict the maintenance cost C for each output restriction condition for each of a plurality of candidate times T _M1 , T _M2 , and T _M3 for the next maintenance of the parts of the wind turbines 100A to 100C.
Note that the maintenance candidate times T _M1 , T _M2 , and T _M3 are preferably set avoiding the period ΔT in which maintenance is difficult to perform. For example, the maintenance candidate times T _M1 , T _M2 , and T _M3 may be set while avoiding the period ΔT in which the probability that access to the windmills 100A to 100C is restricted due to typhoons or aging is high. Since the period ΔT in which maintenance is difficult depends on the location conditions of the wind farm 1, the period ΔT can be known from past weather information and maintenance execution information corresponding thereto. In this way, by setting the maintenance candidate times T _M1 , T _M2 , and T _M3 while avoiding the period ΔT in which maintenance is difficult to perform, an economy that is inevitably generated by waiting until maintenance is possible and refraining from maintenance. Can avoid the loss.

FIG. 6 is a table showing an example of the maintenance cost C predicted for each output restriction condition and maintenance candidate time.
As shown in the figure, when the wind turbine maximum output is not limited (time-dependent change 34A shown in FIG. 5), the maintenance costs C corresponding to the maintenance candidate times T _M1 , T _M2 , and T _M3 are sequentially C ₁ , C _2. , C _full . When the maximum wind turbine output is limited to 90% of the rated output (time-dependent change 34B shown in FIG. 5), the maintenance costs C corresponding to the maintenance candidate times T _M1 , T _M2 , and T _M3 are C ₁ , C ₁ , it is a C _2. Further, when the wind turbine maximum output is limited to 80% of the rated output (time-dependent change 34C shown in FIG. 5), the maintenance costs C corresponding to the maintenance candidate times T _M1 , T _M2 , and T _M3 are 0, C ₁ , C in order. ₁ .

In addition, when predicting the maintenance cost C of the components of each of the wind turbines 100A to 100C, the maintenance cost predicting unit 16 preferably takes into consideration the maintenance cost reduction effect ΔC due to the simultaneous maintenance with other wind turbines.
For example, in the example shown in FIG. 6, when the maintenance of another wind turbine is simultaneously performed at the maintenance candidate time T _M2 , the maintenance cost C is predicted as shown in FIG. 7 in consideration of the maintenance cost reduction effect ΔC. Also good. In the example shown in FIG. 7, the maintenance cost of each output limit conditions in the maintenance candidate time TM ₂ is smaller by ΔC compared to the example shown in FIG.

In order to make the most of the maintenance cost reduction effect ΔC due to the simultaneous maintenance with the other windmills, when the damage or deterioration of the parts of the windmill 100A progresses the fastest, the maintenance candidates for the other windmills 100B and 100C It is preferable that the times T _M1 , T _M2 , and T _M3 include the time when the maintenance of the windmill 100A is expected to be performed immediately before or after the end of the life of the windmill 100A. Alternatively, the state value X of parts of the wind turbine 100A exceeds the threshold value _{X th,} current and other wind turbine 100B if the operation has already been stopped in the wind turbine 100A, 100C maintenance candidate timing _{T M1, T} M2, T It is preferable to include in _M3 .
As a result, the cost reduction effect ΔC due to the simultaneous maintenance with the windmill 100A that is maintained at the earliest time can be reflected in the prediction of the maintenance cost C of the other windmills 100B and 100C.

Further, since the regular maintenance of the wind farm 1 is normally performed for all the windmills 100A to 100C regardless of the damage or deterioration state of the parts, the parts can be maintained and replaced at the same time as the regular maintenance. Maintenance cost reduction effect ΔC can be expected. Therefore, it is preferable that the maintenance candidate times T _M1 , T _M2 , and T _M3 are set so as to include the time for the regular maintenance of the wind farm 1.

Returning to FIG. 1, the power sales revenue prediction unit 18 predicts power sales revenue under a plurality of output restriction conditions for each of the windmills 100 </ b> A to 100 </ b> C. The power sales revenue prediction unit 18 reads, for example, a Weibull distribution representing a wind speed appearance frequency corresponding to each of the windmills 100A to 100C from the wind condition database 19, and predicts a power generation amount under each output restriction condition. You may estimate the electric power sales income obtained from 100A-100C.
The power sales revenue prediction unit 18 may calculate the power sales revenue as zero during the suspension period of the windmills 100A to 100C caused by the total loss of parts. In this case, since the electric power sales revenue obtained from each windmill 100A-100C changes with the maintenance time of each windmill 100A-100C, the electric power sales revenue prediction part 18 sells for every maintenance candidate time T _M1 , T _M2 , T _M3. Electricity revenue may be predicted.

Based on the maintenance cost C of each of the windmills 100A to 100C predicted by the maintenance cost prediction unit 16 and the power sales revenue for each of the windmills 100A to 100C predicted by the power sales revenue prediction unit 18, the profit estimation unit 20 Estimate the profit of the entire farm 1.
As the simplest profit estimation method, the profit estimation unit 20 uses the sum of the power sales revenue of each of the windmills 100A to 100C acquired by the power sales revenue prediction unit 18 to calculate the wind turbines 100A to 100C acquired by the maintenance cost prediction unit 16. By subtracting the sum of the maintenance costs C, the profit of the entire wind farm 1 may be estimated.

  The output restriction condition selection unit 22 selects an output restriction condition for each windmill 100 </ b> A to 100 </ b> C that maximizes the profit of the entire wind farm 1 estimated by the profit estimation unit 20. Based on the selected output restriction condition, a driving command is sent from the driving command unit 24 to each of the wind turbines 100A to 100C. Each of the windmills 100A to 100C performs an operation that satisfies the output restriction condition in accordance with the operation command from the operation command unit 24. For example, when an operation command that restricts the maximum output to 90% of the rated output is sent to the windmill 100A, the windmill controller (not shown) of the windmill 100A causes the power generated by the generator 110 to exceed 90% of the rated output. Do not control.

  Moreover, the maintenance time selection part 23 selects the maintenance candidate time when the profit of the whole wind farm 1 estimated by the profit estimation part 20 becomes the largest about each windmill 100A-100C. The selected maintenance candidate time may be notified to the manager of the wind farm 1 or the worker in charge of the maintenance work by any method. For example, the selected maintenance candidate time may be displayed on the display of the control room of the wind farm 1 or may be displayed on a portable terminal device owned by the worker.

  Next, an operation method of the wind farm 1 will be described. FIG. 8 is a flowchart showing the procedure of the operation method of the wind farm 1.

  As shown in FIG. 8, first, the state monitoring unit 12 of the operation control system 10 acquires the state values X of the components of the windmills 100A to 100C from the wind farm 1 via the signal line 2 (step S2). . Then, the state monitoring unit 12 determines whether or not the state values X of the components of the windmills 100A to 100C are within the management range (step S4). When it is determined that the state value X of the component of each of the windmills 100A to 100C is out of the management range (YES determination in step S4), the process proceeds to step S6, and the windmill having the component is operated from the operation command unit 24. A stop command is sent.

On the other hand, for the wind turbine for which the state value X is determined to be within the management range in step S4, the remaining life prediction of the component is performed by the remaining life prediction unit 14 for each output restriction condition (step S8). At this time, the remaining life prediction unit 14 may perform the remaining life prediction of the part based on the change 30 with time of the state value stored in the state value database 15. For example, the remaining service life prediction unit 14, a change in the state value X at a future current time _{t 1} after curve 32A~32C (see Figure 4) estimated for each output limiting conditions, with these changes curves 32A~32C state time until the value X reaches a threshold _{X th} corresponding to the component total loss (remaining life _{T 100%} of the _{component, T 90%, T 80%} , ...) may be obtained. Further, the remaining service life prediction unit 14, the time until the state value X reaches the threshold _{X th} (remaining life _{T 100%} of the _{component, T 90%, T 80%} , ...) in addition to the greatly affects the maintenance costs The time until the state value X reaches the threshold values X ₁ and X ₂ on the safer side than the threshold value X _th may be obtained.

Thereafter, the process proceeds to step S10, and the maintenance cost predicting unit 16 performs maintenance costs under the respective output restriction conditions for each of the wind turbines 100A to 100C based on the remaining lifetimes T _100% , T _90% , T _80% ,. Predict. At this time, the maintenance cost prediction unit 16 also predicts the maintenance cost C for each output restriction condition for each of the plurality of candidate times T _M1 , T _M2 , and T _M3 for the next maintenance of the parts of the wind turbines 100A to 100C. Good. Further, when predicting the maintenance cost C of the components of each of the windmills 100A to 100C, the maintenance cost reduction effect ΔC due to the simultaneous execution of the periodic maintenance of the wind farm 1 and the simultaneous maintenance of the other windmills is taken into consideration. It is preferable.

In step S12, the power sales revenue prediction unit 18 is used to predict power sales revenue under a plurality of output restriction conditions for each of the windmills 100A to 100C. When predicting the power sales revenue, the power sales revenue during the suspension period of the windmills 100A to 100C caused by the total loss of parts may be calculated as zero. In this case, since the power sales revenue obtained from each of the windmills 100A to 100C varies depending on the maintenance time of each windmill 100A to 100C, even if the power sales revenue is predicted for each maintenance candidate time T _M1 , T _M2 , T _M3. Good.

  Subsequently, the profit estimation unit 20 estimates the profit of the entire wind farm 1 based on the maintenance costs and the power sales revenue of the windmills 100A to 100C predicted in steps S10 and S12 (step S14). For example, as the simplest profit estimation method, by subtracting the sum of the maintenance costs C of the windmills 100A to 100C predicted in step S10 from the sum of the power sales revenue of the windmills 100A to 100C predicted in step S12, The profit of the entire wind farm 1 may be estimated.

  Thereafter, the maintenance time selection unit 23 selects a maintenance candidate time for which the estimated value of the overall profit of the wind farm 1 is maximized for each of the wind turbines 100A to 100C (step S16).

  Further, the output restriction condition selection unit 22 selects an output restriction condition that maximizes the estimated value of the profit of the entire wind farm 1 for each of the wind turbines 100A to 100C (step S18), and performs an operation based on the selected output restriction condition. The wind turbines 100A to 100C are caused to perform the operation command unit 24 (step S20). Thereby, each windmill 100A-100C is drive | operated on the output restriction conditions in which the profit of the whole wind farm 1 becomes the maximum.

  As described above, according to the present embodiment, the power sales revenue and the maintenance cost are predicted for each output restriction condition for each of the windmills 100A to 100C of the wind farm 1, and the output restriction that maximizes the profit obtained from the wind farm 1 is achieved. Since the conditions are selected for each of the windmills 100A to 100C, the profit of the entire wind farm 1 can be effectively improved. That is, instead of focusing on the profits from the windmills 100A to 100C, the operating conditions of the windmills 100A to 100C are determined by focusing on the profits of the wind farm 1 as a whole. Driving becomes possible.

In addition, the output restriction condition and the maintenance candidate cost T _M1 , T _M2 , and T _M3 are predicted for the power sales revenue and the maintenance cost of each of the windmills 100A to 100C, and the output restriction condition that maximizes the profit of the entire wind farm 1 By determining the combination with the maintenance time for each of the windmills 100A to 100C, it is possible to operate the wind farm 1 with further excellent economic efficiency.

  In addition, when the state value X of the part is out of the management range, by stopping the operation of the windmill having the part, the operation of the windmill having the part is stopped before the complete failure of the part, It is possible to prevent a situation in which the windmill is forced to stop for a long period of time.

[Application example of this embodiment]
Next, the economic effect when the above-described embodiment is applied to the wind farm 1 including the windmill A and the windmill B will be specifically described.
FIG. 9 is a diagram illustrating how the output restriction conditions and the maintenance time of the windmill A and the windmill B are determined in the above-described embodiment. FIG. 10 is a diagram showing how the output restriction conditions and the maintenance timing of the wind turbines A and B are determined when the maximization of the profit of each wind turbine is pursued. 9 and 10, when there is no restriction on the maximum output of the windmill A or the windmill B, when the maximum output of the windmill A or the windmill B is limited to 90% of the rated output, and the windmill A or the windmill B The time-dependent change of the maintenance cost C is shown when the maximum output is limited to 80% of the rated output.

As shown in FIG. 10, when the output restriction condition and the maintenance cost are determined so that the profits obtained from the wind turbines A and B are maximized, the output restriction condition of 90% of the rated output and the maintenance time T for the wind turbine A. The combination of _{MA ′} is selected, and for the wind turbine B, the combination of the output restriction condition without restriction and the maintenance time T _{MB ′} is selected. The maintenance times T _{MA ′} and T _{MB ′} are respectively the time immediately before the maintenance cost C of the wind turbine A rises to C _full when the output restriction condition of 90% of the rated output is adopted, and the output restriction condition without restriction. The time immediately before the maintenance cost C of the windmill B at the time of adoption rises to _Cfull is shown.
In the example shown in FIG. 10, the combination of the above-described output restriction condition and maintenance time is selected because the parts are completely lost during the winter season when the wind speed is high and the storm is likely to occur (maintenance difficult period ΔT). This is because the lost profit by stopping, the lost profit by limiting the maximum output of the windmill, and the maintenance cost C are considered, and the profit obtained from each windmill is maximized.

On the other hand, as shown in FIG. 9, when the output restriction condition and the maintenance cost are determined so that the profit obtained from the entire wind farm 1 (wind turbine A and wind turbine B) is maximized, the wind turbine A has 90% of the rated output. the combination selection of the output limiting conditions and maintenance time T _MA, the wind turbine B is a combination of the output limiting conditions and maintenance time T _MB unrestricted is selected. Here, the maintenance times T _MA and T _MB are the time immediately after the maintenance cost C of the wind turbine A rises to C _full when the output restriction condition of 90% of the rated output is adopted, and the output restriction condition without restriction, respectively. maintenance cost C of the wind turbine B indicates when and a C ₂ at the time. In the example shown in FIG. 9, the maintenance timings T _MA and T _MB of the windmill A and the windmill B are substantially matched to enjoy the maintenance cost reduction effect ΔC. Specifically, the maintenance cost reduction effect ΔC can be determined by the labor cost of the worker during the movement between the wind farm 1 and the maintenance base, the movement of the worker between the wind farm 1 and the maintenance base, Charter costs for transportation equipment (trailers, helicopters, crane trucks, crane ships, etc.) for transporting equipment necessary for maintenance can be considered. In the example shown in FIG. 9, the maintenance cost reduction effect ΔC is evenly distributed for the wind turbines A and B.
In the example shown in FIG. 9, the combination of the output restriction condition and the maintenance time that do not necessarily lead to the maximization of the profits of the individual windmills A and B is daringly selected to maximize the profit of the wind farm 1 as a whole. For example, in order to maximize the profit of the windmill A alone, as shown in FIG. 9, the maintenance time _TMA is set immediately before the maintenance cost C rises to _Cfull , and the windmill B is set at the same time. It may be better to perform maintenance. However, in this case, the maintenance time of the windmill B is advanced, and the maintenance of the windmill B is carried out immediately after the maintenance cost C of the windmill B rapidly increases from C ₁ to C _2. The profits of the wind farm 1 as a whole become worse on the contrary.

  As mentioned above, although embodiment of this invention was described in detail, it cannot be overemphasized that this invention is not limited to this, In the range which does not deviate from the summary of this invention, various improvement and deformation | transformation may be performed.

For example, in the above-described embodiment, the case where the maintenance times of the wind turbines 100A to 100C are made different for each part is not considered, but a plurality of candidate maintenance periods are set for each part of the wind turbines 100A to 100C, and the wind farm The combination of the output restriction condition of each of the wind turbines 100A to 100C that maximizes the profit of one whole and the maintenance time for each part of each of the wind turbines 100A to 100C may be determined.
In this case, in predicting the maintenance cost C of each component of the wind turbines 100A to 100C, the cost reduction effect ΔC by performing maintenance in accordance with the maintenance timing of other components belonging to the same wind turbine may be considered. For that purpose, it is preferable that the maintenance candidate time of each part of each of the windmills 100A to 100C is set so as to include the maintenance time of other parts of the same windmill.

1 Wind farm 2 Signal line (bus)
DESCRIPTION OF SYMBOLS 10 Operation control system 12 State monitoring part 14 Remaining life prediction part 15 State value database 16 Maintenance cost prediction part 18 Power sales revenue prediction part 19 Wind condition database 20 Revenue estimation part 22 Output restriction condition selection part 23 Maintenance time selection part 24 Operation command Part 30 State value change with time 32A to 32C Change curve 34A to 34C Maintenance cost 100A to 100C Windmill 102 Blade 104 Hub 105 Rotor 106 Main shaft 107 Main shaft bearing 108 Speed increaser 109 Output shaft 110 Generator 112 Nasser 114 Nasser base plate 116 Tower 118 Yaw slewing bearing 119 Yaw slewing mechanism 120 Hydraulic transmission 122 Hydraulic pump 124 Hydraulic motor 126 High pressure oil line 128 Low pressure oil line

Claims (5)

A method of driving a wind farm with a plurality of windmills,
Predicting the remaining life of parts for each windmill;
For each windmill, predicting power sales revenue under multiple power limiting conditions;
For each windmill, predicting the maintenance cost under each output restriction condition based on the remaining life of the parts;
Selecting, for each wind turbine, an output restriction condition that maximizes the profit obtained from the wind farm, based on at least the power sales revenue predicted for each output restriction condition for each wind turbine and the maintenance cost;
And a step of controlling the operation of each wind turbine based on the selected output restriction condition. The maintenance cost is predicted for each of a plurality of candidate times for the next maintenance of the part,
At least the output restriction condition and the maintenance time at which the profit obtained from the wind farm is maximized based on the power restriction income and the maintenance cost predicted for each wind turbine for the output restriction condition and the candidate time. The wind farm operating method according to claim 1, wherein the wind turbine is determined.   3. The wind farm according to claim 1, wherein in the step of predicting the remaining life, the remaining life of the part is predicted based on a change with time of a state value indicating a damaged state or a deteriorated state of the part. Driving method.   The wind farm according to any one of claims 1 to 3, further comprising a step of stopping the operation of the wind turbine having the part when the state value of the part is out of a management range. how to drive. A wind farm operation control system comprising a plurality of windmills,
A remaining life prediction unit that predicts the remaining life of parts for each wind turbine;
For each windmill, a power sales revenue forecasting unit that predicts power sales revenue under multiple output restriction conditions,
For each windmill, a maintenance cost prediction unit that predicts a maintenance cost under each output restriction condition based on the remaining life of the component,
Output restriction condition selection that selects, for each windmill, an output restriction condition that maximizes the profit obtained from the wind farm based on the power sales revenue and the maintenance cost predicted for each output restriction condition for each windmill. And
An operation control system for a wind farm, comprising: an operation command unit that sends an operation command to each wind turbine based on the selected output restriction condition.

Images